Slurry is any mixture of solid particles and a liquid. Many industrial and mining processes operate on the basis of a wetted product to facilitate chemical reactions. It is also convenient to handle bulk materials in the form of a slurry.
There are many different types of slurry with vastly differing properties. The flow properties are influenced a number of factors including:
Many types of slurry are very abrasive and can be thought of as liquid sandpaper! The abrasive nature of slurries can be reduced by reducing the flow velocity, however many slurries have a minimum flow velocity to avoid deposition of the solid particles.
In comparison to transporting the same mass of dry product, pumping slurry requires the transport of the dry product as well as the carrier liquid. Therefore the required energy is higher than moving just the dry product (However it should be noted that most dry transport mechanisms require high mass carrier systems e.g. rolling stock).
In comparison to pumping water, slurries are usually denser (certainly in mining applications). The higher density results in a higher pressure to pump up a given elevation. Typically the friction pressure losses are also higher so a higher pressure is required to pump at a given velocity over a given distance. Therefore for a given volume a higher amount of work is required to be performed on the slurry in comparison to water.
The most frequent approach to pumping slurry is to make use of centrifugal slurry pumps. A rotating impeller is used to impart kinetic energy to the slurry in the same manner as a conventional water pump. Although this is the most common approach there are a number of serious disadvantages:
The impeller rotates at high speed in the slurry. The relative velocity between the slurry and the impeller is very high resulting in high abrasion rates. In order to reduce the effect of the abrasion the number of impeller vanes is reduced and each vane is thicker in comparison to a water pump. The result is that the slurry impeller is far less efficient than a water impeller. As the pump operates the slurry wears the impeller out, resulting in even lower efficiency and high maintenance costs to replace the impellers periodically. Some impellers are rubber lined which reduces the efficiency further. As a general rule, the larger the diameter of the impeller the higher the efficiency of the slurry pump. See Figure 1 and Figure 2.
As a general process design rule, slurry pumps are designed to be operated to the left of their best efficiency point i.e. slurry pumps are never selected to operate at their best efficiency point.
Slurry pump performance curves are given for pumping pure water. However the presence of solids in the liquid reduces the efficiency and maximum head of the pump in comparison to pumping pure liquid. The solids derating is due to slip between the liquid and the solids as the slurry accelerates into and decelerates out of the impeller. The extent of the solids derating is a function of the:
average particle size
specific gravity of the solids
concentration of solids
A solids derating chart is shown in Figure 6. Typical solids deratings can vary from 0.7 to 0.95.
The maximum pressure rise per pump is 3 – 6 bar depending on the size of the pump. Typical sites require multiple pumps to be placed in series to achieve the required slurry delivery pressure. A typical pump installation therefore requires multiple plinths, interconnecting pipe work, switch gear, gland service water, and an overhead crane to perform maintenance on all the pumps. See Figure 3.
Typically the rotational speed of slurry impellers is adjusted by changing ratios of a V-belt drive. Well maintained V-belt drives can achieve efficiencies of 97%. However in practice such drives are seldom well maintained.
All centrifugal pumps require a seal where the shaft enters the pump casing. In the case of a water pump this is achieved rather simply through the use of a gland or a mechanical seal. In the case of a slurry pump this is far more complex because the slurry erodes the gland. This problem is overcome by injecting high pressure water into the pump to keep the slurry away from the gland. This requires large quantities of clean (usually potable) water at pressures higher than the pump discharge pressure. This wastes huge quantities of potable water, dilutes the slurry, and requires additional power to pump the water into the seal, and then pump the additional water through the slurry pipeline.
Figure 1. - High Efficiency Water Pump Impeller
Figure 2. - Low Efficiency Slurry Pump Impeller
Figure 3. - Multiple Centrifugal Slurry Pumps Placed in Series to Achieve Required Delivery Pressure
Figure 4. - Low Efficiency of Centrifugal Slurry Pumps leads to Large Power Consumption
Figure 5. - Highly Abrasive Environment leads to Frequent Impeller Replacement
The efficiency of the impellor and V-belt drive decrease over time due to wear of the components. The overall efficiency of the pumping system is seldom measured so operators and maintenance staff do not know the actual efficiency. Slurry pump motors are oversized as a standard design practice so that there is additional capacity. As the efficiency of the components drops during operation the excess capacity is utilised without any indication to the operating staff i.e. more power is consumed but the system still pumps so no action is taken.
HR = Head on Slurry / Head on Water
ER = Efficiency on Slurry / Efficiency on Water
CV = % Concentration of Solids in Slurry
by True Volume
d50 = Average Particle Size (mm)
D = Impeller Diameter (mm)
SGS = Specific Gravity of Solids
d50 = 0.365 mm
SGS = 2.65
CV = 30%
D = 365 mm
d50/D = 0.00096
HR ≈ 0.84
ER ≈ 0.80
Figure 6. - Centrifugal Slurry Pump Solids Derating Chart
In addition to the losses described above the pumps are typically driven by electrical motors, usually operating at an efficiency of 95%. All of the losses are cumulative i.e. Overall efficiency = pump efficiency x solids derating x V-belt efficiency x motor efficiency. In addition the gland service water power requirement must be added to the overall energy requirement. The net result is shown in Table 1. An example hydraulic power requirement is also shown, with a typical gland service water power requirement. The System Efficiency shown takes into account the slurry pump losses as well as the gland service water power requirement.
Limited use has also been made of positive displacement slurry pumps for very high pressure applications i.e. greater than 50 bar but typically in the 100 bar range. These pumps are characterised by:
High maintenance costs
Lower power cost due to their high efficiency.
Positive displacement pumps either pump the slurry directly using reciprocating pistons, or separate the slurry from the piston using a diaphragm and oil filled chamber. The piston displaces the oil which in turn displaces the diaphragm. The piston displacement is relatively small in comparison to the flow rate resulting in high cycle rates.
Table 1: Overall Centrifugal Slurry Pump Efficiency
|Parameter||Large, well maintained centrifugal slurry pump||Smaller, poorly maintained centrifugal slurry pump|
|Best Pump Efficiency Point (BEP)||0.8||0.68|
|Efficiency point to left of BEP||0.75||0.64|
|Overall Efficiency (%)||0.62||0.44|
|Example Hydraulic Power Required [kW]||1400||1400|
|Electrical Power Required [kW]||2251||3187|
|Gland Service Water Power [kW]||350||350|
|Net Power [kW]||2601||3537|
|System Efficiency (%)||54||40|
The Phoenix solution to slurry pumping is to pump clean water in a closed loop and transfer the energy from the water to the slurry. The slurry is kept isolated from the water by means of a highly flexible impermeable bladder. The bladder is housed within a pressure vessel. The same water is recycled continuously so apart from the initial water charge, the system does not use water. The water is pumped by a conventional high efficiency water pump (79 – 83% efficiency). There is a minor energy loss within the system but overall the slurry is pumped at just below the efficiency of the water pump. The system can be thought of as a positive displacement pump but with a water piston. Two vessels are operated simultaneously so that the water pump and slurry discharge pressure and flow rate remains constant i.e. one vessel is filling with water while the other is filling with slurry.
The use of a water pump means that:
A very efficient impeller is utilised resulting in a high pump efficiency
V-belt drives are not required so there are no drive losses
Glands are not required with the following advantages:
Gland service water is not required so there is no wastage of potable water
Gland service water pumps are not required
Gland service water pump power is not required
Gland maintenance is eliminated
Solid particles are not accelerated and decelerated through the impeller so there is no solids derating.
The pump can be selected such that it operates at the Best Efficiency Point (BEP).
Very high pumping pressures can be achieved through the use of a single water pump i.e. a single Phoenix Slurry Pump.
Slurry velocities are kept very low throughout the system to minimise the effect of erosion.
The following figures illustrate the operation of the system. In Figure 7 the system is shown in a primed state. Slurry is shown in brown while water is shown in blue. The system is fed gravitationally from a slurry hopper which must be elevated above the system. Slurry flows into each vessel (Vessel A and B) through a slurry inlet non return valve for each vessel. The valves are named Slurry Inlet A (SIA) and Slurry Inlet B (SIB). Slurry outlet valves (SOA and SOB) prevent slurry which has already been discharged from the system flowing back into the vessels. The static head of the slurry forces the water out of the bladders and into a water tank. On each vessel there is a water inlet and outlet valve (WIA, WIB, WOA, WOB). During the prime step WIA and WIB are closed while WOA and WOB are open to allow the water to flow into the water tank. The system is controlled using the water valves. Operation of the water valves controls the flow of the water which in turn controls the flow of slurry. The slurry valves are simple non return valves which only respond to the flow of the slurry i.e. close when the slurry flow stops and open when the slurry flows in a forward direction.
Figure 7. - Phoenix SP P&ID Step 1 Primed
Step 2 is shown in Figure 8. The water pump has been started, WOA and WOB are closed, and WIA is open. This forces water into Vessel A under pressure which inflates the bladder and thereby pressurises the slurry. SIA is closed so the slurry is pumped out SOA and into the slurry discharge line. At the end of the step Vessel A is completely filled with water. Vessel B is still in a primed state.
Figure 8. - Step 2 pumping through Vessel A, B primed
Step 3 is shown in Figure 9. WIB is open so water is pumped into Vessel B, inflating bladder B and pumping slurry out of the vessel. SIB is closed so the slurry is pumped out SOB and into the slurry discharge line. WIA is closed and WOA is opened allowing slurry to flow back into Vessel A and displace the water from the vessel.
Figure 9. - Step 3 pumping through Vessel B, refilling Vessel A with slurry.
Step 4 is shown in Figure 10. Vessel B is being filled with slurry while water is being pumped into Vessel A. At the end of Step 4 the system switches back to Step 3 and continues to loop between Step 3 and 4.
Step 4, the process is repeated by pumping through A and filling B.
In the above explanation the description states that the system switches directly between the two vessels. The actual operation is slightly more complex. The vessel that is having water pumped into it is at the pump discharge pressure, which is the same as the slurry discharge pressure e.g. 16 bar. The vessel filling with slurry is slightly above atmospheric pressure e.g. 1 bar gauge. Switching directly between the two would result in enormous pressure transients in the water pump discharge and slurry discharge. These transients would result in reduced mechanical life of the pump and pipe work, and reduced pump efficiency. To overcome this each vessel is pre pressurised or depressurised prior to switching. This takes a small period of time, and during this time the water pump must continue to discharge water at constant pressure and flow rate. In order to gain this time, the vessel filling with slurry is filled at a greater rate than the other vessel is filled with water. While the one vessel is filled with water, the other vessel is depressurised, filled with slurry, and pre pressurised. In addition if for example water was being pumped into Vessel A, as the switch to B takes place, WIB is opened before WIA is closed. For a short period of time water is filled into both vessels. The above process is referred to as pressure balancing.
The consequence of the time required for pressure balancing is that the slurry inflow rate must be higher than the slurry discharge rate. The slurry discharge is continuous while the slurry inflow is discontinuous i.e. it starts and stops every cycle. This results in the water return to the pump being discontinuous while the pump discharge is continuous. In order to ensure a continuous water flow to the pump the water is buffered using a water tank at atmospheric pressure, called the overlap tank.
The bladder is very flexible and operates within a pressure vessel. During normal operation the bladder is never stretched. The control system must ensure that the inflow of slurry does not push the bladder out the top of the vessel and the inflow of water does not push the bladder out the bottom of the vessel. This is achieved by pumping a fixed quantity of water into the bladder and then allowing the same quantity of water out of the bladder. The water quantity is measured using a flow meter. Every flow meter has a measurement error. Should the effect of the measurement error be cumulative then the bladder may still be pushed out the top or bottom of the vessel over a large number of cycles. To overcome this, the bladder is brought back to a known position on every slurry fill cycle using a component referred to as a valve tube. When the vessel is filled with slurry the valve tube prevents further flow of water out and slurry into the vessel. At this point the water meter is reset to zero. A finite length of time is required for this operation. The time required to depressurise, pre pressurise, and detect that the vessel is full of slurry is called the overlap time. The time from when a vessel is filled with slurry and pre pressurised until the system switches to the vessel is called the overlap margin. The system is always operated with a positive overlap margin. The maximum slurry discharge flow rate is automatically adjusted to keep the overlap margin positive.
During the slurry fill cycle, the slurry inflow accelerates to a peak flow rate, maintains the peak flow rate, and then decelerates as the vessel fills, coming to a complete stop when the vessel is full of slurry. The peak flow rate is a function of the vertical height between the slurry hopper, the vessels and the overlap tank. The deceleration phase is controlled by the control system and the water out valves (WOA and WOB).
The slurry valves simply consist of a polyurethane coated steel ball which seats on an annular valve seat. The valve seat is positioned horizontally, with the flow vertically upwards. The valve seat is designed to minimise deposition on the seat ensuring a drip tight shut off. The ball density is designed to be higher than the slurry. During zero flow conditions the ball settles down onto the valve seat due to its higher density. During forward flow conditions the flow pushes the ball up into the valve housing and bypasses the ball. When the flow stops the ball settles and then prevents reverse flow. Due to the operation of the system, the ball never opens or closes against a differential pressure i.e. the valve is never partially open under flow conditions. This ensures that there are no high velocities which result in accelerated wear.
In summary the following key technologies have allowed the development of the system:
A fixed volume of water is pumped in every cycle and then the same volume is measured out.
The bladder is brought back to a known position every cycle and the flow meters are reset. This eliminates system creep due to measurement errors.
The filling vessel is filled with slurry faster than the other vessel is pumped full of water in order to create overlap time.
The overlap time is created to depressurise, pre pressurise, and detect that the vessel is full of slurry
The overlap time is used to ensure that the water pump discharge and system slurry discharg,br>e is at a constant pressure and flow rate.
Low wear design
Operated such that they are never opened or closed against a differential pressure
Overlap time requires rapid inflow
The inertia of the rapid slurry inflow is carefully damped
The concept of pumping water and then transferring the energy to slurry is not new. Patents for similar concepts date back to 1935. However the development of key technologies described in the previous section were required to produce a commercially viable, robust system.
The ISPS team has been working on various forms of the process since 1998. Initially the system was developed to transfer cold water from surface to underground and use the static head to displace dirty warm water back to surface. The system, known as a water transformer, was operated at 120 bar. The system was then configured for pumping slurry and circa 2007 a slurry pump pilot plant was commissioned. From 2009 the commercialisation process was started with the development of a full size system, installed at Doornkop gold plant. The full size system underwent a design revision to increase the slurry inflow rate in order to increase the slurry discharge flow rate.
A second unit is being assembled for installation at Amandelbult, with a design flow rate of 350 m3/hr at 16 bar.
The pilot plant has been setup in a lab where it is used to demonstrate the system on different slurries and slurry densities. A slurry hopper feeds the unit, which then pumps the slurry back to the hopper through a pressure reducing orifice, simulating the required discharge pressure.
The system can be thought of as a skid mounted process plant used to pump slurry. Figure 11 and Figure 12 show photographs of the first full size unit installed at Doornkop. Figure 13 shows an isometric view of the production unit. The main difference is that on the production unit the pump and motor is located on the frame. This makes the unit more compact and simplifies site installation. The bladder and valve tube removal has also been simplified through the provision of an integral lifting device and the removal of the roof.
The system is skid mounted and completely factory assembled. During transport the system is rotated onto the slurry valve side i.e. the pump and motor is on top. The slurry valves are removed for transport. The stairs, ladders, and platforms are installed on site. The system is shown ready for transport in Figure 14.
There is a lower platform for maintaining the water valves and an upper platform for bladder replacement. The pump, motor, VSD, and slurry valves are maintained from ground level.
The system is controlled by means of a PLC. The system is primed, started, and stopped remotely from the plant control system, or at the unit itself.
Figure 11. - Side view of the first full size unit installed at Doornkop
Figure 12. - Aerial view of the first full size unit installed at Doornkop
Figure 13. - Isometric view of the production unit (first installation at Amandelbult)
Figure14 - System rotated onto its side and loaded for transport
Figure 15 shows some typical test data during 4 cycles. The top graph shows the slurry inlet pressure. There are minor transients as the slurry accelerates into the system and decelerates as the system fills.
The second graph shows the slurry pump discharge pressure. From the graph it can be seen that the pump discharge pressure remains constant, even as the system switches from vessel to vessel. This ensures that the pump has a long service life and operates continuously at peak efficiency.
The third graph shows the slurry discharge pressure. Pressure transients of very short duration and magnitude are evident as the system changes between vessels. However the disturbance is so minor that the discharge can be regarded as constant.
The fourth graph shows the pressure in Vessel A. From the graph the pumping stage, depressurisation, slurry refill, and pre pressurisation are evident.
The fifth graph shows the water in / slurry discharge flow rate. The flow meter gives 1 pulse every 10 litres of water. From the graph it is event that the flow rate remains constant.
The final graph shows the water out / slurry in flow rate. From the graph it can be seen that the flow is discontinuous. The higher pulse density reflects the higher flow rate. The flow deceleration is also evident.
Figure 15. - Typical Test Data
Figure 16 shows the pump, vessel A, B, and slurry discharge pressure. The slurry discharge pressure is 5% lower than the water pump pressure. The hydraulic power imparted to the water by the water pump is the product of the volumetric water flow rate and the increase in pressure. The hydraulic power imparted to the slurry by the water is the product of the volumetric slurry flow rate and the increase in pressure. The efficiency of the energy transfer is the ratio of slurry hydraulic power to water hydraulic power. The water and slurry volumetric flow rates are equal. Therefore the efficiency of the energy transfer is the ratio of the slurry discharge pressure to the water pump discharge pressure i.e. 95%. The loss in pressure is attributed to pressure losses in the valves and pipe work.
Figure 16. - Pump, vessel, and slurry discharge pressure
The overall system efficiency is equal to the energy transfer efficiency x the pump efficiency x the motor efficiency. Note that the solids derating, V-belt drive efficiency, and gland service water power requirement is not applicable. The overall efficiency is shown in Table 2. From the table it can be seen that the overall efficiency is 74%. The electrical power required is calculated for the same hydraulic requirement given in Table 1. The net power and percent saving over the two previous cases is also shown.
Table 2 Phoenix Overall System Efficiency
|Hydraulic Power Required [kW]||1400|
|Electrical Power Required [kW]||1892|
|Gland Service Water||0|
|Net Power [kW]||1892|
|Case A||Case B|
|Net Power Saving [kW]||709||1645|
As mentioned previously the use of a water pump on the Phoenix system means that gland service water is not required. Gland service water is required to be clean. This either means additional filtering of process water or the utilisation of potable water. Once the water is pumped into the gland it enters the slurry where it is contaminated by the slurry. This water can no longer be used as potable water. The gland service water system seldom has measurement instrumentation so the actual utilisation of water is generally unknown. With the Phoenix this water usage is not required.
Furthermore the efficiency of the system is independent of the slurry concentration. This allows the slurry concentration to be increased within the limitations of the pipeline and the gravitational inflow requirement. This increase in concentration has a dramatic decrease in the water requirement of the slurry line
The Phoenix system has the following site requirements:
10 m head on slurry supply hopper i.e. the slurry level should be 10 m above the plinth on which the system is installed. Due to water pump specific NPSH requirements the Phoenix may need to be installed slightly below the overlap tank. In this case the slurry level should be 10 m above the base of the overlap tank. Should the 10 m requirement not be met there will be reduction in the slurry inflow rate and the resultant maximum slurry discharge rate.
Max solids size should not exceed 20 mm in diameter (TBC).
An instrument air supply at 5 bar. A 2.2 kW compressor is the minimum requirement.
Initial charge of potable water. Approximately 15 m3.
Prime / Run / Shutdown signal from the plant control system
There are two drain valves at the bottom of the unit used to drain the slurry side prior to maintenance. Provision needs to be made to drain the discharged slurry into a bund area.
The system is mounted on two plinths. There is a gap between the plinths to facilitate cleaning. The plinths shall have mounting bolts set in position.
The inlet pipe size is 350 NB and the outlet pipe size is 300 NB. For the 16 bar system the flange drilling is as per SABS 1123 1600/3.
Replace coating on valve balls and seats
Service exchange on water valves
Every 5 years
Replace water valve actuators
Every 20 years
Replace pressure vessels i.e. the system is designed for a 20 year life
The maximum flow rate of the system is currently 350 m3/hr. New production units will feature an increased flow rate of 400 m3/hr.
The maximum pressure of the unit is determined by:
The maximum pump pressure
The operating pressure of the
flow meters, and
The current maximum pressure is 16 bar. 30 and 66 bar vessels have been designed and water components sourced. The remaining item to be designed for higher pressures is the slurry valves.
As with all new technology there is increased risk in comparison to existing technology. The risk has been minimised through a long and detailed development process with continual testing of systems.
However the overall risk of any site needs to be considered. The continual use of inefficient systems which consume large quantities of energy and water place entire sites at far greater risk in an environment where there is increasing pressure on power and water resources.
Slurry pumping is a very energy intensive process which has traditionally been performed using centrifugal slurry pumps, which pump at low efficiency, waste water, and are maintenance intensive. The Phoenix Slurry Pump has been developed to pump high volumes of slurry at high pressure with a very high overall system efficiency with a low maintenance requirement. Slurry pumping is achieved by pumping water using an efficient water pump and then transferring the energy to the slurry using a bladder and vessel arrangement. Two vessels are utilised to achieve a constant water pump and slurry discharge pressure and flow rate. The system is operated in a manner that ensures smooth switching from vessel to vessel.
The net result is:
High efficiency slurry pumping through:
Utilisation of a very efficient pump operating at its BEP
No solids derating
Elimination of V belt drives
Elimination of gland service water pump power
Glands are not required with the following advantages:
No wastage of potable water
Gland service water pumps and piping are not required
Gland maintenance is eliminated
Very little maintenance
Only one pump is required to achieve the required discharge pressure
Switchgear for one pump instead of multiple pumps
A very low life cycle cost